Capítulo 4 Regulación del sector
4.2. Otras regulaciones internacionales
3.4.1
Stock returns analysis
The results of the stock market reaction after downgrades are shown in Table 3.4. The [0; 1] event window is statistically significant for downgrades with an ACAR of −0.72%. Highly significant abnormal return can be measured only on the first trading day after the release. Weak significant effects can be observed on the event day 0. Rating announcements released after the close of the exchange are priced in the stock price with a lag of one day. During the [−2; 2] event window, ACAR is−0.51%, but not significant. During the investigation period, 104 downgrade reviews are observed, and the results for downgrade reviews are shown in Panel B.
Downgrade reviews have a stronger impact on the stock market. Although Germany has a strong bank-customer relationship, rating reviews are new and relevant information for equity investors. The abnormal returns for rating reviews are−1.56% during the [−2; 2] event window. All event windows that include the release day are statistically highly significant. The [−2;−1] event window suggests that there is only small leakage prior to the announcement. Bannier and Hirsch (2010) find statistically highly significant negative effects for direct and watch-preceded U.S. downgrades in the [−1; 1] window of −1.89% and −3.10%, respectively. Jorion and Zhang (2007b) even indicate CARs of−4.43% at the 1% level of significance for rating downgrades in the United States. Norden and Weber (2004) also find highly statistically significant and negative effects for their worldwide data set. Our analysis shows that market reaction to rating downgrades is considerably smaller in Germany. Therefore, the results for rating downgrades support the hypothesis that rating changes are less important to equity investors in Germany than in the
Table 3.4: Stock market reactions on downgrade actions and downgrade reviews;∗,∗∗,∗∗∗denote significance at the 10, 5 and 1% level, respectively.
Event Median t-test SIGN BMP-test Corrado-test Sample window ACAR CAR (t-value) (Z-Score) (Z-score) (Z-scroe) size Panel A: Average abnormal stock returns by rating downgrade
[−2;−1] 0.32% 0.13% 1.04 −1.26 1.29 1.48 166 [−1;0] −0.09% −0.04% −0.32 −0.06 −0.25 0.21 166 {0} −0.31% −0.21% −1.88∗ −1.84∗ −1.89∗ −1.56 166 [0;1] −0.72% −0.72% −2.71∗∗∗ −2.90∗∗∗ −2.61∗∗∗ −2.62∗∗∗ 166 [−1;1] −0.50% −0.26% −1.44 −1.22 −1.31 −1.08 166 [−2;2] −0.51% −0.20% −1.04 −0.85 −0.95 −1.39 166 Panel B: Average abnormal stock returns by downgrade review
[−2;−1] −0.55% −0.44% −1.29 −1.64 −1.75∗ −2.05∗∗ 104 [−1;0] −0.92% −0.92% −2.33∗∗ −3.24∗∗∗ −2.65∗∗∗ −3.49∗∗∗ 104 {0} −0.60% −0.54% −2.06∗∗ −3.05∗∗∗ −2.23∗∗∗ −3.24∗∗∗ 104 [0;1] −1.23% −1.30% −3.14∗∗∗ −3.52∗∗∗ −3.30∗∗∗ −4.52∗∗∗ 104 [−1;1] −1.55% −1.66% −3.63∗∗∗ −4.02∗∗∗ −3.85∗∗∗ −4.67∗∗∗ 104 [−2;2] −1.56% −1.73% −2.82∗∗∗ −3.54∗∗∗ −3.24∗∗∗ −4.22∗∗∗ 104 Panel C: Differences of downgrades and downgrade reviews
Event D_Median two-sample Wilxocon rank-sum window D_ACAR CAR t-test (t-value) test (Z-score) [−2;−1] 0.87% 0.57% 1.69∗ −2.05∗∗ [−1;0] 0.82% 0.89% 1.73∗ −2.90∗∗∗ {0} 0.29% 0.34% 0.92 −1.80∗ [0;1] 0.51% 0.58% 1.11 −1.51 [−1;1] 1.04% 1.40% 1.88∗ −2.63∗∗∗ [−2;2] 1.06% 1.53% 1.40 −2.50∗∗
United States. Rating reviews are significant, because rating reviews reveal new information to the market.
The analysis is repeated for upgrades and upgrade reviews. Panel A of Table 3.5 shows the results for rating upgrades. The ACARs are negative and non-significant. These results suggest that upgrade information is not relevant to the stock market. Panel B reports results for upgrade reviews. Contrary to downgrade reviews, there are no statistically significant effects on upgrade reviews. The ACAR ranges from a positive value of 0.61% during the [0; 1] event window to 1.21% during the event windows [−1; 1] and [−2; 2]. Overall, rating upgrades do not seem to be relevant information. The findings confirm prior results for U.S. firms (Goh and Ederington, 1993, 1999) and European banks (Gropp and Richards, 2001). Upgrade announcements are not relevant in market-based or bank-based systems.
Downgrade announcements are less relevant in bank-based systems. Rating announcements reveal new information to the market, so we find significant effects for rating reviews. Positive rating announcements show no significant effects.
3.4.2
Reason given in rating announcements
Prior studies provide evidence that the reason given by agencies for their rating actions is relevant for the stock market reaction (Goh and Ederington, 1993, 1999; Gropp and Richards, 2001). Table 3.4 presents the results for rating downgrades according to the categories defined earlier, operating performance and capital structure. The results support our prior results of rating downgrades.
Panel A shows the results for operating performance rating downgrades. ACAR is slightly lower in comparison with capital structure downgrades, and is significant at the 5% level of significance only during the [0; 1] event window.
Capital structure downgrades (Panel B) have no significant effect. During the [−2; 2] event window, the overall ACAR for capital structure downgrades is positive. We find results similar to those of Gropp and Richards (2001) for rating downgrades. Capital structure has a positive effect on stock prices during the [−2; 2] event window, whereas operating performance is negative. These
Table 3.5: Stock market reactions on upgrade actions and upgrade reviews; ∗ , ∗∗ , ∗∗∗ denote significance at the 10, 5 and 1% level, respectively.
Event Median t-test SIGN BMP-test Corrado-test Sample window ACAR CAR (t-value) (Z-Score) (Z-score) (Z-scroe) size Panel A: Average abnormal stock returns by rating upgrade
[−2;−1] 0.19% 0.18% 0.96 −1.02 1.21 1.02 105 [−1;0] −0.05% −0.20% −0.27 −0.72 −0.14 −0.12 105 {0} −0.03% −0.20% −0.25 −0.42 −0.17 −0.30 105 [0;1] −0.13% −0.26% −0.71 −1.24 −0.59 −0.56 105 [−1;1] −0.15% −0.19% −0.67 −0.86 −0.50 −0.38 105 [−2;2] 0.09% −0.43% 0.25 −0.60 0.33 −0.04 105 Panel B: Average abnormal stock returns by upgrade review
[−2;−1] 0.40% 0.81% 0.47 −0.41 0.84 1.11 18 [−1;0] 0.51% 0.23% 1.00 −0.94 1.31 0.81 18 {0} −0.09% −0.27% −0.24 −0.24 −0.21 −0.01 18 [0;1] 0.61% 0.37% 0.74 −0.54 0.56 0.32 18 [−1;1] 1.21% 0.37% 1.57 −1.33 1.70∗ 0.93 18 [−2;2] 1.21% 2.16% 1.42 −1.50 1.49 1.44 18 Panel C: Differences of upgrades and upgrade reviews
Event D_Median two-sample Wilxocon rank-sum window D_ACAR CAR t-test (t-value) test (Z-score) [−2;−1] −0.21% −0.62% −0.34 0.54 [−1;0] −0.56% −0.43% −1.09 0.94 {0} 0.06% 0.07% 0.19 −0.10 [0;1] −0.74% −0.63% −1.32 1.06 [−1;1] −1.36% −0.57% −2.14∗∗ 1.68∗ [−2;2] −1.13% −2.59% −1.23 1.65
results are not significant, but the difference the downgrade categories is weakly significant during the [−2; 2] event window.
Our prior findings show that rating upgrades have no significant effect on the German stock market. This result is confirmed when the upgrades are divide into the two categories of rationales (Table 3.7). The value of the ACARs is very similar and not significantly different from zero. During the [0; 1] event window, the average abnormal stock return due to an operating performance rating change is−0.21% and−0.19% for a capital structure change. In line with Bannier and Hirsch (2010) and Holthausen and Leftwich (1986), rating upgrades have no significant effect on the stock market.